DESCRIPTION: This proposal addresses the mechanism of mismatch repair in yeast and the role of this process in genome maintenence. The focus is on the MSH2 gene product (Msh2) an analogue of E. coli mutS. Dr. Alani and co-workers demonstrated that like mutS, Msh2 is directly involved in binding to mismatched bases. They also demonstrated interaction between Msh2 and Mlh1 and Pms1, the yeast homologues of MutL, another component of the E. coli mismatch repair system. A mutagenesis of the MSH2 gene is proposed with the goal of identifying alleles that differentially affect aspects of mismatch repair. The initial assay is based on the mutator phenotype of msh2 mutants (10-30X for CanR and 1-700 for a lacZ dinucleotide repeat reversion assay). Conditional and dominant mutants are sought as well as mutations that are dominant when over expressed (GAL10::MSH2). A suppressor analysis of these mutants is also described with the goal of further defining domains of Msh2 involved in specific activities and other proteins with which Msh2 might interact. A biochemical characterization of Msh2 binding to oligonucleotides carrying mismatches is proposed to define parameters that could be compared to the mutant proteins. By analogy to MutS and Msh1 (the yeast mitochondrial homologue) an ATPase activity for Msh2 will be investigated. Mutant Msh2 proteins will be compared to wildtype for binding to mismatched oligonucleotides, binding to Pms1 and Mlh1, ATPase activity with the goal of defining domains responsible for these activities. Association between Msh2 and known recombination proteins will be monitored both by affinity chromatography, immunoprecipitation and co-immunolocalization. The immunolocalization approach will be extended to meiotic cells. Alani also addresses the proposal that Msh2 has a role in restricting the formation of strand exchange intermediates between homologous DNAs by terminating or reversing strand exchange in the vicinity of mismatched bases. A related suggestion is that mismatch detection is a central component of the mechanisms that inhibit crossover events between homeologous sequences, hence prohibiting translocations and deletions that would occur by crossover between ectopic sites. Further tests to determine the role of mismatch detection in restricting homeologous recombination are described. Alani will determine the consequences of mismatch repair defects on the efficiency of homeologous gene conversion and the proportion of such events associated with crossovers using an assay from Jim Haber's laboratory that detects homeologous recombination events and distinguished those associated with crossovers by a simple replicating test. A key component of the ability of E. coli to correct mismatched bases derived from replication errors is its ability to detect the old strand by the presence of hemimethylated sites. The mutH protein is responsible for coupling mismatch detection to strand discrimination. Yeast do not have methylated DNA and no homologue of mutH has been detected. Alani proposes to determine the functions involved in strand discrimination as coupled to mismatch correction by a genetic screen for synthetic lethality patterned after an observation from E. coli. recA- dam- cells are inviable but mutS- recA- dam- cells are viable. One interpretation is that the mismatch repair, in the absence of strand discrimination provided by the dam system, makes nicks on both strandswhich become DSBs and are dependent on recombination for their repair. Starting with a rad52 mutant carrying a msh2 conditional allele, colonies propagated at the nonpermissive condition for msh2 will be screened for the inability to grow at the permissive condition for msh2. Among the synthetic lethal mutations from this screen, Alani hopes to identify mutations in genes involved in strand discrimination.